From LifeWiki
Jump to navigation Jump to search

I'm Tanner Jacobi, another Life enthusiast. I've contributed to a few things: improving the synthesis for the dart, and then progressively greater roles in the first syntheses for the crab, 30P5H2V0, weekender, Pushalong 1, x66, and B29. I've also improved syntheses for rare natural objects by looking at apgsearch results, notably the fore and back and Eater 2. I also discovered the syringe. So, just a bunch of odds and ends.

To-do in no particular order

  • Find a 3/8 orthogonal spaceship, or at least some promising partials.
  • Collect H-to-block converters, like the H-to-G collection.
  • Finish synthesizing the "skewship" (x = 14, y = 9, rule = B3/S23 b3o$o2bo6b3o$o4bo4bo2bo$3bobo2bo4bo$2o4bobobo$5bobo4b2o$3bo4bo$4bo5bo$9bo!)
  • Formalize an expansion and codification for still life nomenclature.
  • Continue developing a genetic spaceship search algorithm with a smooth fitness landscape--crossing raw soups isn't working, because the fitness landscape is very rough. Current idea is to allow for "fall" cells that break the rules of B3/S23 after X generations, and award increased fitness when fall cells are changed into real spaceship parts. See the Eltrut problem video; it demonstrates that algorithms can be changed to turn a rough fitness landscape into smooth one, which deals with the issues presented here.